6502 02 58 f3 case doesn't match Visual 6502

[tom-seddon]

{ "name": "02 35 44", "initial": { "pc": 1837, "s": 197, "a": 33, "x": 239, "y": 61, "p": 108, "ram": [ [1837, 2], [1838, 53], [1839, 68]]}, "final": { "pc": 1837, "s": 197, "a": 33, "x": 239, "y": 61, "p": 108, "ram": [ [1837, 2], [1838, 53], [1839, 68]]}, "cycles": [ [1837, 2, "read"], [1838, 53, "read"], [1838, 53, "read"]] }

It suggests there should be a read from the PC on the 3rd cycle, but the bus activity in Visual 6502 shows a read from $ffff:

cycle	ab	db	rw	Fetch	pc	a	x	y	s	p
0	0000	02	1	unknown	0000	aa	00	00	fd	nv‑BdIZc
0	0000	02	1	unknown	0000	aa	00	00	fd	nv‑BdIZc
1	0001	58	1		0001	aa	00	00	fd	nv‑BdIZc
1	0001	58	1		0001	aa	00	00	fd	nv‑BdIZc
2	ffff	00	1		0002	aa	00	00	fd	nv‑BdIZc
2	ffff	00	1		0002	aa	00	00	fd	nv‑BdIZc

Then 2 reads from fffe, and then endless ffff.

cycle	ab	db	rw	Fetch	pc	a	x	y	s	p
3	fffe	00	1		0002	aa	00	00	fd	nv‑BdIZc
3	fffe	00	1		0002	aa	00	00	fd	nv‑BdIZc
4	fffe	00	1		0002	aa	00	00	fd	nv‑BdIZc
4	fffe	00	1		0002	aa	00	00	fd	nv‑BdIZc
5	ffff	00	1		0002	aa	00	00	fd	nv‑BdIZc
5	ffff	00	1		0002	aa	00	00	fd	nv‑BdIZc
6	ffff	00	1		0002	aa	00	00	fd	nv‑BdIZc
6	ffff	00	1		0002	aa	00	00	fd	nv‑BdIZc
7	ffff	00	1		0002	aa	00	00	fd	nv‑BdIZc
7	ffff	00	1		0002	aa	00	00	fd	nv‑BdIZc
8	ffff	00	1		0002	aa	00	00	fd	nv‑BdIZc
8	ffff	00	1		0002	aa	00	00	fd	nv‑BdIZc
9	ffff	00	1		0002	aa	00	00	fd	nv‑BdIZc
9	ffff	00	1		0002	aa	00	00	fd	nv‑BdIZc
10	ffff	00	1		0002	aa	00	00	fd	nv‑BdIZc
10	ffff	00	1		0002	aa	00	00	fd	nv‑BdIZc
11	ffff	00	1		0002	aa	00	00	fd	nv‑BdIZc
11	ffff	00	1		0002	aa	00	00	fd	nv‑BdIZc
12	ffff	00	1		0002	aa	00	00	fd	nv‑BdIZc
12	ffff	00	1		0002	aa	00	00	fd	nv‑BdIZc
13	ffff	00	1		0002	aa	00	00	fd	nv‑BdIZc
13	ffff	00	1		0002	aa	00	00	fd	nv‑BdIZc
14	ffff	00	1		0002	aa	00	00	fd	nv‑BdIZc
```

[tom-seddon]

A longer trace with the $02 at $0180 - suggesting the program counter doesn't affect the addresses read.

cycle	ab	db	rw	Fetch	pc	a	x	y	s	p
0	0000	a9	1	LDA #	0000	aa	00	00	fd	nv‑BdIZc
0	0000	a9	1	LDA #	0000	aa	00	00	fd	nv‑BdIZc
1	0001	00	1		0001	aa	00	00	fd	nv‑BdIZc
1	0001	00	1		0001	aa	00	00	fd	nv‑BdIZc
2	0002	4c	1	JMP Abs	0002	00	00	00	fd	nv‑BdIZc
2	0002	4c	1	JMP Abs	0002	00	00	00	fd	nv‑BdIZc
3	0003	80	1		0003	00	00	00	fd	nv‑BdIZc
3	0003	80	1		0003	00	00	00	fd	nv‑BdIZc
4	0004	01	1		0004	00	00	00	fd	nv‑BdIZc
4	0004	01	1		0004	00	00	00	fd	nv‑BdIZc
5	0180	02	1	unknown	0180	00	00	00	fd	nv‑BdIZc
5	0180	02	1	unknown	0180	00	00	00	fd	nv‑BdIZc
6	0181	58	1		0181	00	00	00	fd	nv‑BdIZc
6	0181	58	1		0181	00	00	00	fd	nv‑BdIZc
7	ffff	00	1		0182	00	00	00	fd	nv‑BdIZc
7	ffff	00	1		0182	00	00	00	fd	nv‑BdIZc
8	fffe	00	1		0182	00	00	00	fd	nv‑BdIZc
8	fffe	00	1		0182	00	00	00	fd	nv‑BdIZc
9	fffe	00	1		0182	00	00	00	fd	nv‑BdIZc
9	fffe	00	1		0182	00	00	00	fd	nv‑BdIZc
10	ffff	00	1		0182	00	00	00	fd	nv‑BdIZc
10	ffff	00	1		0182	00	00	00	fd	nv‑BdIZc
11	ffff	00	1		0182	00	00	00	fd	nv‑BdIZc
11	ffff	00	1		0182	00	00	00	fd	nv‑BdIZc
12	ffff	00	1		0182	00	00	00	fd	nv‑BdIZc
12	ffff	00	1		0182	00	00	00	fd	nv‑BdIZc
13	ffff	00	1		0182	00	00	00	fd	nv‑BdIZc
13	ffff	00	1		0182	00	00	00	fd	nv‑BdIZc
14	ffff	00	1		0182	00	00	00	fd	nv‑BdIZc

[fuel-pcbox]

on an original MOS 6502, the 02 opcode is KIL, which glitches the CPU and forces a hard reset. That's why Visual 6502 is showing a difference.

[tom-seddon]

My real hardware is packed away but I'm 99% certain there is no reset when executing 02. The CPU appears to get stuck.

I'd expect visual6502 to be accurate, so my assumption is that this is a bug in the test suite?